Group Shift Pointwise Convolution for Volumetric Medical Image Segmentation

TL;DR

This paper introduces Group Shift Pointwise Convolution (GSP-Conv), a novel method that simplifies 3D convolutions to reduce complexity while maintaining or improving segmentation performance in volumetric medical images.

Contribution

The paper proposes GSP-Conv, a new approach combining pointwise convolutions with a parameter-free Group Shift operation to enhance efficiency and effectiveness in 3D medical image segmentation.

Findings

Achieves comparable or better segmentation accuracy than traditional 3D convolutions.

Reduces model parameters and FLOPs by up to 27 times.

Demonstrates effectiveness on PROMISE12 and BraTS18 datasets.

Abstract

Recent studies have witnessed the effectiveness of 3D convolutions on segmenting volumetric medical images. Compared with the 2D counterparts, 3D convolutions can capture the spatial context in three dimensions. Nevertheless, models employing 3D convolutions introduce more trainable parameters and are more computationally complex, which may lead easily to model overfitting especially for medical applications with limited available training data. This paper aims to improve the effectiveness and efficiency of 3D convolutions by introducing a novel Group Shift Pointwise Convolution (GSP-Conv). GSP-Conv simplifies 3D convolutions into pointwise ones with 1x1x1 kernels, which dramatically reduces the number of model parameters and FLOPs (e.g. 27x fewer than 3D convolutions with 3x3x3 kernels). Na\"ive pointwise convolutions with limited receptive fields cannot make full use of the spatial image context. To address this problem, we propose a parameter-free operation, Group Shift (GS), which shifts the feature maps along with different spatial directions in an elegant way. With GS, pointwise convolutions can access features from different spatial locations, and the limited receptive fields of pointwise convolutions can be compensated. We evaluate the proposed methods on two datasets, PROMISE12 and BraTS18. Results show that our method, with substantially decreased model complexity, achieves comparable or even better performance than models employing 3D convolutions.